Orientation Detection for 2D Vessel Segmentation for Angiography-FFR

Abstract

An apparatus for evaluating a vessel of interest and a corresponding method are provided, wherein the modeling of hemodynamic parameters using a fluid dynamic model can be verified by deriving feature values ​​from the segmented vessel of interest and inputting these feature values ​​into a classifier. The classifier can then determine based on the feature values ​​whether the segmentation has been performed from proximal to distal, from distal to proximal or cannot be determined from the provided data. An incorrect segmentation order can thus be identified and possibly corrected, thereby avoiding inaccurate simulation results.

Description

Technical Field

[0001] The present invention relates to an apparatus, a corresponding method and a computer program for evaluating vessels of interest. In particular, the present invention relates to an apparatus employing a trained classifier to automatically detect and / or correct vessel segmentation orientations used for generating vessel segmentations of physiological models including fluid dynamic models. Background Art

[0002] Accurate evaluation of coronary artery disease is extremely important for determining the best treatment options. Such evaluation requires an understanding of the hemodynamics in coronary artery physiology. For this purpose, metrics such as fractional flow reserve (FFR) or instantaneous wave-free ratio (iFR) can be used to accurately assess a patient's coronary artery disease. Both FFR and iFR are metrics for the pressure drop of blood along the vessel of interest, for example, caused by a stenosis in the vessel of interest. It can be determined as the ratio of the pressure (Pd) distal to the stenosis to the pressure (Pa) in the aorta.

[0003] Therefore, FFR measurements must be performed during maximum blood flow (ie, under hyperemia), which may cause discomfort to the patient. In contrast, iFR measurements are performed at rest during a specific period in diastole, thereby avoiding the need to induce hyperemia in the patient.

[0004] In the past, FFR and / or iFR measurements were usually performed invasively, by measuring pressure at locations distal to and adjacent to the lesion. In recent years, efforts have been made to non-invasively determine FFR and / or iFR values ​​by means of so-called "virtual" approaches. According to the virtual approach, the fluid dynamics in the coronary arteries of a patient are simulated based on a physiological model including a fluid dynamics model representing the blood flow through one or more vessels of interest.

[0005] Therefore, the physiological model is generated based on the diagnostic image data of the coronary vasculature. That is, the vessel of interest as represented by the image data is segmented and the physiological model is generated based on the segmentation. In order to perform vessel segmentation, the user usually manually indicates the start position and the end position of the vessel of interest to be segmented. Therefore, the start position is considered to correspond to the proximal end and the end position is considered to correspond to the distal end of the vessel segment to be considered. The start position and the end position define the segmentation order of the vessel segmentation from the proximal side to the distal side.

[0006] The correct choice of the start and end positions of the vessel to be segmented is therefore important for achieving accurate fluid dynamics estimation using the fluid dynamics model, since this choice determines the orientation of the vessel segment, in other words, the "vessel segmentation orientation".

[0007] Typically, the vessel segmentation is oriented such that blood is assumed to flow from a starting position to an ending position. Thus, fluid dynamic modeling is performed such that the starting position of the vessel segmentation corresponds to a proximal position and the ending point corresponds to a distal position.

[0008] On the other hand, if the start and end positions are interchanged, for example during manual selection, the vessel segmentation orientation will be reversed, corresponding to an incorrect blood flow direction from a distal position to a proximal position. In this case, fluid dynamics modeling will provide inaccurate or incorrect results. Therefore, the process is quite error-prone. Summary of the invention

[0009] It is therefore an object of the present invention to prevent incorrect evaluation of fluid dynamics in a vessel of interest using a physiological model comprising a fluid dynamic model of said vessel of interest.

[0010] More specifically, an object of the present invention is to provide an apparatus and method for automatically detecting a vessel segmentation orientation of a vessel of interest to be segmented for generating a physiological model including a fluid dynamics model and for verifying the vessel segmentation orientation so that a user can be aware of incorrectly evaluated vessels of interest.

[0011] This object is achieved by a device for evaluating a patient's vessel of interest, the device comprising: an input unit configured to receive segmentation data including an indication of at least one vascular segment obtained by segmenting the vessel of interest; a calculation unit configured to derive at least one vascular parameter for the at least one vascular segment; and a classification unit configured to receive the at least one vascular parameter and classify the at least one vascular parameter to obtain an indication of the segmentation orientation of the at least one vascular segment, wherein the classification unit has been trained using a training data set, the training data set being in the form of a data set associating a set of known vascular segmentation orientations with corresponding values ​​of the at least one vascular parameter.

[0012] In this context, the term segmented data may specifically refer to a data set comprising one or more segments of the vessel of interest (i.e. a vessel that should be considered, for example, to determine the hemodynamic properties of the vessel of interest). The segments may thus be obtained by acquiring at least one diagnostic image using a medical imaging modality and by identifying the vessel of interest in the at least one diagnostic image. The identified vessel of interest is then segmented into the one or more segments. The segment size may thus be predetermined manually by a user or automatically by the device.

[0013] The term vessel parameters may specifically refer to parameters that may assist in deriving additional information about the vessel and / or segmentation. As an example, after segmentation, characteristic values, such as geometric parameter values, are derived for at least part of the vessel segments, for example, corresponding cross-sectional values ​​may be determined for each vessel segment. These cross-sectional values ​​may then be considered as a function of the vessel length in order to determine a route indicating the development of the vessel cross-sectional area from the start position to the end position of the vessel segmentation.

[0014] Therefore, the determined vascular parameters can then be input into the classification unit. The classification unit is provided so that it has been trained using ground truth using a training data set. The training data set can specifically have been trained using ground truth that allows the classification unit to classify one or more input vascular parameters such as to detect deviations in the orientation of vascular segmentation. That is, in the example mentioned above, the classification unit has been trained using ground truth that allows the classification unit to determine the segmentation orientation of the vascular segmentation according to the route of the cross-sectional values ​​as a function of the vascular length. The classification unit can therefore determine that the orientation is the same as assumed, that is: the proximal position of the vascular of interest has been used as the starting position for vascular segmentation, and the distal position of the vascular of interest has been used as the end position for vascular segmentation. The classification unit can also determine that the orientation deviates from the assumption. In some embodiments, this means that the classification unit can determine that the distal position of the vascular of interest has been used as the starting position for vascular segmentation, and the proximal position of the vascular of interest has been used as the end position. In some embodiments, the deviation from the initial assumption may also correspond to the classification unit determining that it is not able to detect which position of the vessel of interest corresponds to the start position and the end position of the segmentation, respectively, and therefore the orientation is unclear.

[0015] Thus, based on the classification, the classification unit may obtain an indication of the orientation of the vessel segmentation. This indication may then be output to the user and / or may be used by the apparatus to automatically correct the segmentation order.

[0016] In some embodiments, the apparatus further comprises a segmentation unit, the segmentation unit being configured to: receive user input indicating a segmentation start position and a segmentation end position for a vessel of interest represented in at least one diagnostic image, and segment the vessel of interest represented in the at least one diagnostic image to obtain the segmentation data, the segmentation data comprising a plurality of segments, each segment being obtained for a particular position among a plurality of positions along the length of the vessel of interest, wherein a first position corresponds to the segmentation start position and a last position corresponds to the segmentation end position. In some embodiments, the at least one diagnostic image has been acquired using X-ray angiography.

[0017] The apparatus may further comprise a segmentation unit configured to segment at least one diagnostic image to generate the segmentation data and to provide the segmentation data to the input unit. In this context, it should be understood that the segmentation may be based on one or more diagnostic images representing at least the vessel of interest.

[0018] The one or more diagnostic images may have been acquired using a medical imaging modality. Any imaging modality may be used as long as it allows identifying the vessel of interest and performing vessel segmentation. As examples of suitable medical imaging modalities, computed tomography, magnetic resonance imaging, ultrasound imaging, etc. are mentioned.

[0019] In some embodiments, the at least one diagnostic image may specifically include a two-dimensional X-ray angiography image. For this purpose, a contrast agent may be introduced into the vessel of interest. Subsequently, an X-ray angiography of the contrast-filled vessel is performed. Thus, the acquired image may specifically have a low degree of perspective contraction and low overlap. Furthermore, the contrast filling should be sufficient to provide a suitable contrast. After acquisition, the X-ray angiography image may then be presented to the user for manual segmentation input.

[0020] In some embodiments, the user may be specifically requested to manually input a start position and an end position for the segmentation. That is, typically, the at least one diagnostic image is presented to the user who may consider the vessel and provide input indicating points that the user considers to correspond to the start position and the end position.

[0021] In some embodiments, the device automatically assigns a start position and an end position for segmentation. This automatic assignment may also be based on classification. More specifically, the automatic assignment may be performed by training a (separate) classifier in the device with a training data set of diagnostic images that allows the classifier to determine potential start and end positions for the data.

[0022] After selecting the starting position and the ending position, the segmentation is started. Thus, the segmentation proceeds along the length of the vessel of interest from the starting position to the ending position. That is, the segmentation of the vessel of interest is performed at a plurality of points along the length of the vessel of interest. Thus, the starting position corresponds to a first position of the plurality of positions, and the ending position corresponds to a last position of the plurality of positions. Furthermore, a corresponding segment is obtained at each position of the plurality of positions.

[0023] Thus, the obtained plurality of segments - one for each of the positions - are included in the segmented data. In this context, it should be understood that the term position should be interpreted broadly, as referring to a segment of finite length that has been obtained at a specific point along the length of the vessel of interest, i.e. along the longitudinal direction inside the vessel of interest.

[0024] Subsequently, at least one vascular parameter is calculated based on multiple segments. That is, for each segment from the multiple segments, features, such as geometric parameter values, can be obtained. The eigenvalue can then be considered as a function of multiple positions along the length of the vascular of interest for further analysis. In some embodiments, the eigenvalue is considered as a function of the multiple positions and can be calculated by the computing unit, and the resulting vascular parameters (such as the slope of the function) can be input into the classification unit for classification. In some embodiments, the eigenvalue itself and its corresponding position and / or the index for which its corresponding segment has been derived can be inserted into the classification unit, and the classification unit itself then determines the vascular parameters and classifies the vascular parameters to obtain an indication.

[0025] According to some embodiments, the at least one vascular parameter is based on one or more of: a geometric parameter value for each of the plurality of segments, and / or a first value relating to the first position and the second position, respectively, and / or a second value relating to the first position and the second position, respectively, and / or an acquisition parameter value relating to an acquisition setting used to acquire the at least one diagnostic image, and / or a dynamic parameter derived from the at least one diagnostic image.

[0026] In some embodiments, the at least one vascular parameter can be derived by analyzing a plurality of characteristic values ​​obtained for the plurality of segments. One such characteristic value can specifically involve deriving a geometric parameter value for each segment in the segment, such as a cross-sectional value or a diameter determined for each segment in the segment. The geometric parameter value can then be regarded as a function of a plurality of positions along the vascular of interest or a function of an index of a corresponding segment for which a geometric parameter value has been determined. Subsequently, the slope of the function can be derived. In some embodiments, the slope can specifically be determined by a linear regression through all geometric parameter values. The vascular parameter to be input into the classification unit then includes the slope determined as described above.

[0027] Alternatively or additionally, the at least one vascular parameter can be derived by superimposing at least one (two-dimensional) diagnostic image with a two-dimensional coordinate system and taking into account the x-coordinates, y-coordinates for the start point and the end point (indicated by the start position and the end position) of the vascular segment. That is, as corresponding feature values, a second value x1 for the start point at the start position can be determined, and a second value x2 for the end point at the end position can be determined. The vascular parameters to be input into the classification unit then include an x ​​difference value x12 indicating the difference (in the x-direction) between the two second values ​​x1, x2.

[0028] Alternatively or additionally, as corresponding feature values, a first value y1 for a starting point at the starting position may be determined, and a first value y2 for an ending point at the ending position may be determined. The vascular parameters to be input into the classification unit then include a y difference value y12 indicating the difference (in the y direction) between the two first values ​​y1, y2.

[0029] Alternatively or additionally, information independent of the segmentation and the diagnostic image may be used for the at least one vascular parameter. In some embodiments, one or more acquisition parameters indicative of acquisition settings may be provided to the classifier, such as a C-arm angle, patient orientation information and / or positioning of a table on which the patient is located. In some embodiments, the tracked position of an intravascular measurement device (such as a guide catheter) as visible in the X-ray image may also be used.

[0030] Alternatively or additionally, a dynamic parameter, which is a parameter indicative of the flow dynamics through the vessel of interest, may be derived from the at least one diagnostic image. More specifically, one or more diagnostic images may be used to retrieve additional information about the dynamics of fluid flow, in particular contrast agent flow, over time. In some embodiments, a time series of diagnostic images may be used to track the flow of contrast agent. This allows the flow direction through the vessel of interest to be determined directly.

[0031] The slope value and / or the x-difference value and / or the y-difference value and / or one or more acquisition parameter values ​​and / or the kinetic parameters may then be provided to the classification unit individually or in combination. The classification unit may be trained using ground truth for associating one or more of the vascular parameters (such as the slope value, the x-difference value, the y-difference value, one or more acquisition parameter values ​​and / or the kinetic parameters) with the vascular segmentation orientation.

[0032] In this regard, the training dataset may be provided in the form of a dataset of vascular segmentation data with known orientations including the parameters listed above. In other words, the training dataset associates a set of known vascular segmentation orientations with corresponding values ​​of at least one vascular parameter.

[0033] The classification unit can then use the trained ground truth to detect whether the vessel segmentation orientation is "normal" (ie, proximal to distal), "inverted" (ie, distal to proximal), or unclear. Based on this detection, the classification unit can provide an indication of the classification result.

[0034] According to some embodiments, the classification unit may implement a plurality of subunits. Thus, the subunits may be specifically trained using different training data sets. For example, each subunit may be trained using a different training data set, wherein the vessel segmentation orientation is associated with a corresponding subset of the vessel parameters.

[0035] As an example, one of the subunits in the classification unit has received a training data set of vessel segmentation data with known orientation including vessel parameters slope values, x-difference, y-difference, C-arm angle indication as acquisition parameters and contrast agent flow information as kinetic parameter. An additional subunit has received a training data set of vessel segmentation data with known orientation including slope values, x-difference, y-difference and contrast agent flow information as kinetic parameter. In this case, if the above parameters are available as input, the first subunit can be used, and if the C-arm angle is missing, the second subunit can be used.

[0036] In some embodiments, some classification subunits may have been trained to use a single vascular parameter for segmentation orientation determination. In some embodiments, the classification subunits may have been trained to use two, three, four, etc. vascular parameters in different combinations.

[0037] Such an arrangement allows the device to be adapted to different situations where different inputs (in particular, different subsets of vascular parameters) can or will be provided to determine the orientation of the vessel segmentation.

[0038] In some embodiments, the classification unit comprises a support vector machine.

[0039] The term support vector machine can be understood in a conventional manner, that is, as referring to a binary linear classifier that implements a supervised, semi-supervised or unsupervised learning model. The classifier is trained using a training data set that includes data labeled as belonging to one of two categories. The training output is a model that represents the values ​​of the training data set as points in the mapped space so that the input is divided into one of the two categories with a clear gap as wide as possible between them.

[0040] Each new input to be classified is then mapped into the same space predicted to belong to one of two categories depending on the side of the gap on which they are provided. In the present case, the two categories include orientations that are proximal to distal (i.e., "correct") and orientations that are distal to proximal (i.e., "incorrect"). Furthermore, any results provided inside the gap are not clearly classified, and as such, are considered "unclear" in the present case.

[0041] In this regard, it should be understood that in some embodiments, the classification unit and / or the calculation unit can be implemented using a neural network that has been trained using a corresponding training data set as described above. The neural network can receive the segmentation data and can determine the orientation of the segmentation based on the training.

[0042] According to some embodiments, the apparatus further comprises a display unit, wherein the display unit is configured to generate a graphical representation in response to the indication of the segmentation orientation, and to display the graphical representation to a user.

[0043] In some embodiments, the device may be specifically provided with a display unit, which is configured to generate a graphical representation in response to the indication of the segmentation orientation. The graphical representation can then be displayed to the user. The graphical representation may include a representation of the vessel of interest and corresponding markings indicating the start position and end position of the vessel segmentation. The graphical representation may specifically include a graphical representation of the indication. In some embodiments, the graphical representation may include a warning message that the orientation is in the wrong order (i.e., from distal to proximal) or that the orientation cannot be determined (i.e., unclear). In some embodiments, the graphical representation may indicate that the orientation is in the correct order.

[0044] According to an embodiment, the indication of the segmentation orientation comprises one of: an error indication indicating that the segmentation orientation is from distal to proximal, a verification indication indicating that the segmentation orientation is from proximal to distal, or a clarity indication indicating that the segmentation orientation is unclear.

[0045] The term indication may particularly refer to a visual, audible and / or tactile indication which may be used to inform a user of the vessel orientation of the segmentation orientation.

[0046] The error indication may thus specifically correspond to an indication that the segmentation has been performed in the wrong order (i.e., the vessel segmentation orientation is reversed, from distal to proximal) such that the resulting modeling of the fluid dynamics in the segmented vessel of interest may be inaccurate. The error indication may be visual and / or auditory, and may specifically be in the form of a visual and / or auditory warning message. The error indication may also be configured to terminate the fluid modeling process so that no incorrect results are achieved.

[0047] The verification indication may specifically correspond to an indication that the segmentation has been performed in the correct order (i.e., the vessel segmentation orientation is correct, from proximal to distal). Therefore, the resulting modeling of the fluid dynamics in the segmented vessel of interest may be considered accurate. The verification indication may also be visual and / or auditory. In some embodiments, the verification indication may be such that no indication is output at all when the orientation is found to be in the correct order.

[0048] The clarity indication may correspond to an indication that the order of segmentation cannot be determined (i.e., is unclear). That is, the resulting modeling of the fluid dynamics may be accurate (if the order of segmentation is already correct) or inaccurate (if the order is already incorrect), but it is not possible to determine which is the case. The clarity indication may be visual and / or auditory, and may specifically include a warning message. In some embodiments, the visual and / or auditory warning message may be different from the visual and / or auditory warning message used for the error indication. In some embodiments, the clarity indication may cause the fluid dynamics model to stop modeling the fluid dynamics through the vessel of interest.

[0049] In some embodiments, the apparatus comprises a correction unit configured to perform a correction segmentation in response to the indication of the segmentation orientation of the vessel segmentation comprising the error indication.

[0050] According to some embodiments, the apparatus may also be configured to perform segmentation correction in response to the error indication. That is, the apparatus determines that the segmentation has been performed in the wrong order, and may switch the start and end positions for the segmentation in response to the determination. The segmentation may then be redone in the correct order, and subsequent modeling of the fluid dynamics may be performed based on the new segmentation.

[0051] In some embodiments, the apparatus further comprises an output unit which outputs said indication of said segmentation orientation of said vessel segmentation.

[0052] According to some embodiments, the device may also be provided with an output unit for providing visual, auditory and / or tactile output to the user. In some embodiments, the output unit may specifically include a sound emitter to output an auditory signal that warns the user that the orientation should not be used or verifies to the user that the orientation can be used. In some embodiments, the output unit may be part of or correspond to the display unit. In some embodiments, the output unit may be separate from the display unit, and the display unit may only be used to interact with the user for defining the start and end positions of the segmentation and representing the modeling results.

[0053] In some embodiments, the classification unit is trained using a training dataset comprising a plurality of known segmentation orientations.

[0054] The classification unit may be trained by inputting a training data set. The training data set may specifically include a plurality of vessel segmentations with known orientations. That is, the classification unit receives information about which vessel segmentation has which orientation. Based on this, the classification unit may be trained specifically about which vessel parameter value may indicate which vessel segmentation orientation. In the case where the classification unit comprises a support vector machine, the unclear segmentation orientation may be defined by a distance metric in the feature space and a predetermined minimum threshold for the distance.

[0055] According to another aspect, a method for evaluating a vessel of interest of a patient is provided, the method comprising the following steps: receiving segmentation data comprising an indication of at least one vascular segment resulting from vascular segmentation of the vessel of interest, deriving at least one vascular parameter for the at least one vascular segment, and classifying the at least one vascular parameter to obtain an indication of a vascular segmentation orientation of the at least one vascular segment, wherein the classification is performed based on ground truth derived from a training data set associating a set of known vascular segmentation orientations with corresponding values ​​of the at least one vascular parameter.

[0056] In some embodiments, the method further comprises the step of outputting the indication of the segmentation orientation of the vessel segmentation. In some embodiments, the method further comprises the step of correcting the segmentation orientation of the vessel segmentation if the segmentation orientation is from distal to proximal.

[0057] In a further aspect, a computer program for controlling an apparatus according to the invention is provided, said computer program being adapted to perform the method steps according to the invention when run by a processing unit. In yet another aspect, a computer readable medium having stored thereon the computer program listed above is provided.

[0058] It shall be understood that the apparatus of claim 1, the method of claim 12, the computer program of claim 14 and the computer readable medium of claim 15 have similar and / or identical preferred embodiments as particularly defined in the dependent claims.

[0059] It shall be understood that a preferred embodiment of the present invention can also be the independent claim or any combination of the above embodiments with the corresponding dependent claims.

[0060] These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In the following figures:

[0062] Figure 1 An apparatus for evaluating a vessel of interest according to an embodiment is schematically illustrated.

[0063] Figure 2 An example classification result according to an embodiment is illustrated.

[0064] Figure 3 An exemplary graphical representation to be presented on a display unit is illustrated.

[0065] Figure 4 A flow chart showing a method for evaluating a vessel of interest according to an embodiment. DETAILED DESCRIPTION

[0066] The illustrations in the drawings are schematic. In different drawings, similar or identical elements are provided with the same reference numerals.

[0067] Figure 1 An exemplary embodiment of an apparatus for evaluating a vessel of interest in a patient is schematically shown. Figure 1 In an exemplary embodiment of , the apparatus 1 comprises: an input unit 100 comprising a segmentation unit 101 , a calculation unit 200 comprising a modeling unit 201 , a classification unit 300 , a correction unit 400 , and a display unit 500 comprising an output unit 501 .

[0068] In accordance with Figure 1In an exemplary embodiment of the apparatus 1, the apparatus 1 receives a diagnostic image 10 from an imaging modality via an input unit 100. In this specific example, the diagnostic image 10 corresponds to a two-dimensional X-ray angiography image 10 of a vessel of interest that has been filled with a contrast agent. The angiography image 10 may have a low degree of perspective contraction, low overlap, and good contrast agent filling. The input unit 100 provides the angiography image 10 to the display unit 500. The display unit 500 generates a graphical representation of the angiography image 10 and outputs the graphical representation to a user. The display unit 500 further generates a graphical representation for prompting the user to provide a request indicating a start position and an end position in the vessel of interest for performing segmentation. After receiving the user input regarding the start position and the end position, the segmentation unit 101 may then perform segmentation of the angiography image 10.

[0069] Therefore, the generated segmentation data 11 may then be provided to the calculation unit 200 and the modeling unit 201 .

[0070] The modeling unit 201 can use the segmentation data 11 to provide a physiological model of the vessel of interest, the physiological model including a fluid dynamics model representing the fluid dynamics through the vessel of interest. Therefore, the generated model can be provided to the display unit 500. The display unit 500 can generate a graphical representation of the model of the vessel of interest and display the representation to the user.

[0071] In accordance with Figure 1 The fluid dynamics model used in the exemplary embodiment of can specifically correspond to a lumped parameter fluid dynamics model. In such a lumped parameter fluid dynamics model, the fluid dynamics of the vessel is approximated by the topological structure of discrete entities. As an example, a vascular system (such as a vascular tree) can be represented by a topological structure of resistor elements that all have a specific resistance. Therefore, the outlet at the far end of the vessel is also represented by a specific resistor element. The resistor element is then connected to the ground, such as to represent the terminal of the vessel. Similarly, the corresponding resistor element can be connected to a series of resistor elements representing the vessel of interest, such as to represent the outflow from the vessel of interest at a specific bifurcation. These resistor elements can also be connected to the ground generally.

[0072] The lumped parameter fluid dynamics model reduces the number of dimensions compared to other approaches such as Navier-Stokes etc. Therefore, using a lumped parameter fluid dynamics model may allow for a simplified calculation of the fluid dynamics inside the vessel and is therefore particularly beneficial in terms of processing time.

[0073] The physiological model may also include a geometric model of the vessel of interest. Figure 1In the specific embodiment, since the input is already a single two-dimensional X-ray angiography image 10, the geometric model may specifically correspond to a two-dimensional or quasi-three-dimensional geometric model in which a third dimension is approximated.

[0074] The computing unit 200 may use the segmented data 11 to compute one or more values ​​indicative of vascular parameters. Figure 1 In an exemplary embodiment of , the calculation unit 200 can determine the diameter of the vessel of interest specifically for each segment in the segmented data 11, and the diameter derived therefrom can then be considered as a function of the number of segments. This allows the derivation of a linear slope for the vessel diameter along the length of the segmented vessel of interest. This slope can then be used as the first vessel parameter.

[0075] Furthermore, the calculation unit 200 may consider the vessel in an xy coordinate system. Subsequently, the calculation unit may determine the xy coordinates of the vessel segmentation start position and the vessel segmentation end position. The calculation unit 200 may then be configured to determine a first value indicating the difference of the y coordinates for the start position and the end position, respectively. Furthermore, the calculation unit 200 may be configured to determine a second value indicating the difference of the x coordinates for the start position and the end position, respectively.

[0076] In accordance with Figure 1 In an exemplary embodiment of , the xy coordinates for the segmentation start position may be (1, 2), and the xy coordinates for the segmentation end position may be (3, 5). The first value may thus correspond to a value of 3, and the second value may correspond to a value of 2. The first value may be considered a second vascular parameter. The second value may be considered a third vascular parameter.

[0077] The three vascular parameters, namely the slope of the diameter and the different values ​​along the x-direction and the y-direction, may then be provided from the calculation unit 200 to the classification unit 300 .

[0078] In accordance with Figure 1 In an exemplary embodiment of the present invention, the classification unit 300 includes a linear support vector machine that has been trained on a data set of vascular segmentation data with known orientations. Based on this training, the classification unit 300 can be configured to classify the segmentation data based on the derived vascular parameters. The classification can thus be specifically performed into one of three categories. The first category can correspond to vascular segmentations that have been performed in the correct order (i.e., from proximal to distal). The second category can correspond to vascular segmentations that have been performed in an incorrect order (i.e., from distal to proximal). The third category can correspond to vascular segmentations whose order is unknown or unclear.

[0079] For this, if the classification unit includes a linear support vector machine, Figure 2Corresponding to an exemplary output of the classification unit 300. Schematic diagram 30 shows the use of i c The distance d in the feature space of the index used for segmentation c . Thus, the results are separated by a separation plane indexed by dc=0.0. The small dots on the left hand side 32 of the schematic diagram 30 correspond to segmentations with an orientation from distal to proximal (i.e., they have been segmented in the wrong order). The small dots on the right hand side 33 of the schematic diagram 30 correspond to segmentations with an orientation from proximal to distal (i.e., they have been segmented in the correct order). The gray area 31 indicates a predetermined range for situations where the segmentation orientation is considered unknown.

[0080] In other words: according to Figure 2 In an exemplary embodiment, the distance in the feature space (in the range from -1.5 to +1.5) is a measure of the segmentation orientation, wherein a distance in the feature space between -1.5 and -0.1 corresponds to a distal to proximal orientation, a distance in the feature space between 0.1 and 1.5 corresponds to a proximal to distal orientation, and a distance in the feature space between -0.1 and 0.1 corresponds to an "unclear" orientation.

[0081] Back to Figure 1 In the embodiment of the present invention, the classification result is then evaluated by the device 1 .

[0082] In accordance with Figure 1 In a specific embodiment of the present invention, this means that if the classification unit 300 determines that the orientation is from proximal to distal, the fluid dynamics model can be used to determine the fluid dynamics through the vessel of interest. In this case, the classification unit 300 can provide a command to the modeling unit 201 to perform fluid dynamics modeling and output the results to the display unit 500. In addition, the classification unit 300 can provide a verification indication to the display unit 500, prompting the display unit 500 to generate a graphical representation thereof to confirm to the user that the segmentation has been performed in the correct order. The display unit 500 can also draw a graphical representation of the hemodynamic parameters simulated using the fluid dynamics model.

[0083] In some embodiments, the fluid dynamics model may have been used to determine the hemodynamics before the classification results are available. In these cases, the classification unit may cause the modeling unit 501 to transmit the results to the display unit 500 for graphical representation. In some embodiments, the modeling unit 501 may also transmit the results before receiving a command from the classification unit, and the classification unit may only provide a verification indication to the display unit 500 to ensure verification by the user.

[0084] If the classification unit 300 determines that the orientation is from the distal side to the proximal side, an error indication may be generated by the classification unit 300 and provided to the display unit and / or the output unit for outputting a warning to the user. The warning may correspond to, for example, Figure 3 A graphical representation of a warning message 51 is shown in . Alternatively or additionally, the warning may correspond to an audible or tactile indication to the user.

[0085] In this case, the classification unit 300 may also be configured to instruct the modeling unit 201 not to generate and / or output a model of the vessel of interest. Alternatively or additionally, the modeling unit 201 may have generated a physiological model including the fluid dynamics model. In this case, the model may also be transmitted to the display unit 500, and a graphical representation thereof may be generated by the display unit 500 and shown on the display unit 500, whereby the graphical representation may also include a warning message.

[0086] In accordance with Figure 1 In an exemplary embodiment of the present invention, the classification unit 300 may also be configured to transmit the diagnostic image 10 and / or the segmentation data 11 to the correction unit 400. The correction unit 400 may then correct the segmentation of the diagnostic image 10 or the corrected segmentation data 11 to generate segmentation data 11' indicating a correct segmentation order.

[0087] It should be understood that although the correction unit 400 is Figure 1 10 is represented as a separate entity, but it can also be integrated into the segmentation unit 101, whereby the correction unit 300 can specifically correct the start position and the end position, and cause the segmentation unit 101 to re-segment using the corrected start position and the end position to retrieve the segmentation data 11'.

[0088] The segmentation data 11' may then be provided to the modeling unit 201. The modeling unit 201 may use the segmentation data 11' to generate a physiological model including a fluid dynamics model, and use the fluid dynamics model to simulate corresponding hemodynamic parameters. The modeling unit 201 may also provide the simulation results and / or the physiological model to the display unit 500 for graphical representation. The graphical representation may include an indication that the segmentation used for modeling has been automatically corrected.

[0089] The classification unit 300 may determine that the orientation of the segment is unknown. Figure 2 This is the case for all dots within the region 31. Here, the minimum threshold for distance in the feature space (indicated by the grey region 31) has been determined, and each distance value within this threshold region corresponds to an unknown orientation.

[0090] In this regard, it should be understood that the modeling unit 201 may use the segmented data 11 with unknown orientation to generate a physiological model including a fluid dynamics model and simulate hemodynamic parameters.

[0091] However, the classification unit 300 may provide a clarity indication to the display unit 500. The display unit 500 may then generate a graphical representation of the clarity indication, in particular in the form of a warning message to be shown on the display unit 500. Furthermore, the output unit 501 may generate an auditory warning message to indicate to the user that the orientation is unknown, so that the user may review this information and possibly correct the segmentation and / or not rely on possibly inaccurate simulation results.

[0092] Figure 4 The method for evaluating a vessel of interest is schematically illustrated. In step S101, the input unit 100 receives a diagnostic image 10, and Figure 4 In the embodiment of FIG. 5 , the diagnostic image 10 corresponds to a two-dimensional X-ray angiography image 10 of the vessel of interest. In step S102 , the input unit 100 provides the angiography image 10 to the display unit 500 .

[0093] In step S501, the display unit 500 generates a graphical representation of the angiographic image 10. Furthermore, in step S502, the display unit 500 presents the graphical representation to the user and requests the user to indicate a start position and an end position for performing segmentation of the vessel of interest. In step S503, the user provides a corresponding input indicating the start position and the end position. Then, the method proceeds to step S103, in which the segmentation unit 101 performs segmentation of the angiographic image 10.

[0094] In step S104, the segmentation data 11 is provided from the input unit 100 and the segmentation unit 101 to the calculation unit 200 and the modeling unit 201. In step S201, the modeling unit 201 generates a physiological model including a geometric model of the vessel of interest and a fluid dynamics model for simulating fluid dynamics through the vessel of interest using the segmentation data 11, and provides the model to the display unit 500 in step S202. In step S504, the display unit 500 generates a graphical representation of the geometric model and the fluid dynamics model, and displays the representation to the user.

[0095] In addition, the calculation unit 200 uses the segmentation data 11 to calculate one or more values ​​indicating a vascular parameter. Specifically, in step S202, the calculation unit 200 determines the diameter of the vascular of interest for each stage in the segmentation data 11, and determines the slope of the function representing the diameter according to the number of segments. In addition, in step S203, the calculation unit 200 considers the vascular of interest in a two-dimensional (x, y) coordinate system, and determines a first value indicating the difference of the y coordinates of the start position and the end position of the segmentation of the vascular of interest, respectively. In addition, in step S204, the calculation unit 200 determines a second value indicating the difference of the x coordinates of the start position and the end position, respectively. The first value and the second value are then used as a second vascular parameter and a third vascular parameter. In step S205, the calculation unit 200 provides the three vascular parameters to the classification unit 300.

[0096] In step S301, the classification unit 300 classifies the segmented data into one of three categories based on the derived vascular parameters, namely: "proximal to distal" (segmentation in the correct order), "distal to proximal" (segmentation in an incorrect order), or "unclear" (segmentation orientation cannot be determined). In step S302, the classification unit 300 outputs the classification result as one of a verification indication, an error indication, or a clarity indication to the display unit 500. Figure 4 In an exemplary method, if the classification result corresponds to an error indication, in step S302, the error indication is further provided to the correction unit 400.

[0097] In step S505, the display unit 500 generates a graphical representation of the classification result. That is, if the classification result includes a verification indication, the display unit 500 generates a graphical representation of the verification indication and displays it side by side with the graphical representation of the physiological model. Therefore, the graphical representation may generally include a geometric model of the vessel of interest, one or more hemodynamic parameters determined using the fluid dynamics model, and a verification indication indicating that the segmentation orientation has been verified (such as a text message indicating "segmentation orientation from proximal to distal" on the display unit 500), etc. In some embodiments, step S505 may also include outputting an auditory or tactile confirmation signal through the output unit 501.

[0098] If the classification result includes an error indication, in step S505, the display unit 500 generates a graphical representation of the error indication and displays it to the user. In some embodiments, the graphical representation may include a warning message. In some embodiments, the warning message may state "the segmentation orientation is from the distal side to the proximal side" or the like. In some embodiments, the graphical representation may also include a diagnostic image 10 superimposed with the warning message. In some embodiments, the graphical representation may also include a request to prompt the user to manually correct the start position and the end position (i.e., the segmentation orientation). In addition, step S505 may also include outputting an auditory or tactile error signal through the output unit 501.

[0099] In accordance with Figure 4 In the exemplary embodiment of , in step S505, a warning message is displayed on the display unit 500 in the case of an error indication. Then, in step S506, the warning message advances to a message "Correct Orientation" indicating that the correction unit 400 is performing correction of the segmentation.

[0100] In step S401, in response to the error indication, the correction unit 400 retrieves the diagnostic image 10 or the segmentation data 11, and performs correction of the segmentation so that the segmentation is performed in the correct order. In step S402, the correction unit 400 outputs the corrected segmentation data 11' to the modeling unit 201, which then proceeds to generate a physiological model including a fluid dynamics model based on the corrected segmentation data 11' in step S201, and provides the thus corrected model to the display unit 500 in step S504. In step S507, the display unit then displays a graphical representation of the corrected model and a message indicating that the segmentation orientation on which the model is based has been corrected (such as a message stating "Correction model ended. Show results"). In addition, the graphical representation generated and presented in step S507 includes one or more hemodynamic parameters that have been calculated using the fluid dynamics model.

[0101] If in step S302 the classification unit determines that the orientation of the segmentation is unknown, then in step S505 a graphical representation of a warning message indicating that the orientation is unknown is generated. The warning message may, for example, state "ambiguous orientation. Please identify the near end of the segmentation". Furthermore, step S505 may include outputting an auditory or tactile signal via the output unit 501 indicating that the orientation is unclear. The tactile or auditory signal should be different from the tactile or auditory signal used in the case of a warning message in the case of an erroneous indication.

[0102] In some embodiments, the warning message may be displayed on the display unit 500 alongside the diagnostic image 10. In this case, the user may manually input the proximal (and / or distal) ends for segmentation. In other embodiments, the warning message may be displayed alongside the (preliminary and possibly inaccurate) physiological model and the corresponding hemodynamic parameters, and the user may decide whether he or she will rely on the determined parameters or whether he or she prefers to repeat the segmentation and modeling process.

[0103] Although in the embodiments listed above, diagnostic images have been acquired using X-ray angiography, it should be understood that other imaging modalities may similarly be used, such as computed tomography, ultrasound imaging, magnetic resonance imaging, etc. It should also be understood that although the embodiments listed above foresee the use of a single image, multiple images may also be used to generate the model.

[0104] Furthermore, although in the above embodiments, the vessel diameter slope and the difference along the x- and y-directions have been used as vascular parameters for classifying the segmentation, it should be understood that additional features (such as C-arm projection angles, virtual FFR calculations along two directions (i.e., from proximal to distal and from distal to proximal) , The sequence of interventional projections and / or automatically tracked contrast catheter position) can be used as input for the evaluation of the vessel of interest.

[0105] Furthermore, although in the above-listed embodiments, a support vector machine is used as a classifier, it should be understood that other classifiers may also be used. Similarly, the classification unit may implement a neural network for classification of orientations.

[0106] Although in the embodiments listed above the blood flow reserve fraction and / or the instantaneous wave-free ratio have been indicated as hemodynamic indicators, it should be understood that other indicators can also be derived, such as the pressure ratio between the pressure distal to the stenosis and the pressure in the aorta, the coronary blood flow reserve and other flow-, pressure-, etc.-related indicators.

[0107] Furthermore, although in the embodiments listed above, the apparatus itself determines the segmentation data of the diagnostic image data, it should be understood that the term determining is to be interpreted broadly and may also include input of the segmentation data from another apparatus.

[0108] Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims.

[0109] In the claims, the word "comprising" does not exclude other elements or steps, and the word "a" or "an" does not exclude a plurality.

[0110] A single unit or device may fulfill the functions of several items recited in the claims.The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

[0111] Processes like segmentation of at least one diagnostic image, derivation of at least one vascular parameter, classification of at least one vascular parameter, generation of a physiological model, correction of the physiological model and / or segmentation data, etc., performed by one or several units or devices, can be performed by any other number of units or devices. These processes according to the invention can therefore be implemented as program code units of a computer program and / or as dedicated hardware.

[0112] The computer program may be stored / distributed on a suitable medium, such as an optical storage medium or solid-state medium supplied with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems.

[0113] Any reference signs in the claims should not be construed as limiting the scope.

[0114] The present invention relates to an apparatus for evaluating a vessel of interest of a patient, comprising: an input unit configured to receive segmentation data indicating a vessel segmentation of the vessel of interest; a calculation unit configured to derive at least one vessel parameter based on the segmentation data; a classification unit configured to receive the at least one vessel parameter and classify the at least one vessel parameter to obtain an indication of the segmented orientation of the vessel segment. Furthermore, the present invention relates to a method performed by the apparatus.

Claims

1. An apparatus for evaluating a vessel of interest of a patient, the apparatus comprising: an input unit configured to receive at least one diagnostic image representing the vessel of interest, the input unit comprising a segmentation unit configured to receive a user input indicating a segmentation start position and a segmentation end position for the vessel of interest represented in the at least one diagnostic image, and segment the vessel of interest to obtain segmentation data, the segmentation data comprising a plurality of segments, each segment being obtained for a specific position among a plurality of positions along the length of the vessel of interest, wherein a first position corresponds to the segmentation start position and a last position corresponds to the segmentation end position, a modeling unit configured to provide a physiological model using the segmentation data obtained based on the segmentation start position and the segmentation end position indicated by the user input, a computing unit configured to derive at least one vascular parameter for the at least one vascular segment, and a classification unit configured to receive the at least one vascular parameter and classify the at least one vascular parameter to obtain an indication of a vascular segmentation orientation of the at least one vascular segment determined based on the segmentation start position and the segmentation end position indicated by the user input, comprising one of the following: an error indication indicating that the segmentation orientation is reversed, a verification indication indicating that the segmentation orientation is normal, or a clarity indicator indicating that the segmentation orientation is unclear, Therein, the classification unit has been trained using a training dataset, wherein the training dataset is in the form of a dataset associating a set of known vessel segmentation orientations with corresponding values ​​of the at least one vessel parameter.

2. The device according to claim 1, wherein: The at least one diagnostic image has been acquired using X-ray angiography.

3. The device according to any one of claims 1 and 2, wherein: The at least one vascular parameter is based on one or more of: a geometric parameter value for each segment of the plurality of segments, and / or a first value associated with the first position and the second position, respectively, and / or a second value associated with the first position and the second position, respectively, and / or acquisition parameter values ​​associated with an acquisition setting used to acquire the at least one diagnostic image, and / or A kinetic parameter is derived from the at least one diagnostic image.

4. The device according to any one of claims 1 to 2, wherein: The classification unit comprises a plurality of sub-units, each sub-unit having been trained using a different training data set associating the vessel segmentation orientations with a corresponding subset of the at least one vessel parameter.

5. The device according to any one of claims 1 to 2, wherein: The classification unit includes a support vector machine.

6. The device according to any one of claims 1 to 2, further comprising a display unit, wherein: The display unit is configured to generate a graphical representation in response to the indication of the segmentation orientation and to display the graphical representation to a user. 7 . The apparatus according to claim 1 , further comprising a correction unit configured to perform a correction segmentation having a normal segmentation orientation in response to the indication of the segmentation orientation of the vessel segmentation comprising the error indication.

8. The apparatus according to any one of claims 1 to 2, further comprising an output unit which outputs the indication of the segmentation orientation of the vessel segmentation.

9. A method for evaluating a vessel of interest in a patient, the method comprising the steps of: receiving at least one diagnostic image representing the vessel of interest, receiving a user input indicating a segmentation start position and a segmentation end position for the vessel of interest represented in at least one diagnostic image, segmenting the vessel of interest to obtain segmentation data, the segmentation data comprising a plurality of segments, each segment being obtained for a specific position of a plurality of positions along the length of the vessel of interest, wherein a first position corresponds to the segmentation start position and a last position corresponds to the segmentation end position, providing a physiological model using the segmentation data obtained based on the segmentation start position and the segmentation end position indicated by the user input, deriving at least one vascular parameter for the at least one vascular segment, and Classifying the at least one vascular parameter to obtain an indication of a vascular segmentation orientation of the at least one vascular segment determined based on the segmentation start position and the segmentation end position indicated by the user input comprises one of the following: an error indication indicating that the segmentation orientation is reversed, a verification indication indicating that the segmentation orientation is normal, or A clarity indication indicating that the segmentation orientation is unclear, wherein the classification is performed based on ground truth derived from a training data set associating a set of known vascular segmentation orientations with corresponding values ​​of the at least one vascular parameter.

10. The method of claim 9, further comprising the step of outputting the indication of the segmentation orientation of the vessel segmentation.

11. A computer program product comprising a computer program adapted to perform the method according to any of claims 9 and 10 for controlling an apparatus according to any of claims 1 to 8 when the computer program is run by a processing unit.

12. A computer readable medium having stored thereon a computer program adapted to perform the method according to any one of claims 9 and 10 for controlling an apparatus according to any one of claims 1 to 8 when the computer program is run by a processing unit.

Images